Graduation Research Theme List

kayakstarsΤεχνίτη Νοημοσύνη και Ρομποτική

15 Νοε 2013 (πριν από 3 χρόνια και 4 μήνες)

149 εμφανίσεις

(Form
-
1)

Graduation Research Theme List


Date:

15
/

1
1

/ 2009


Supervisor Name

Theme Titles

Maximum number of
students

Dr Ha Quoc Trung

(1)

Enterprise Application Integration research and implementation in
University Environment

02


02

02

(2)

Distributed replication fo
r web application

(3)
Distributed Software Repository System Using P2P model

Dr. Ngo Hong Son

(1)
Development

of

location based
-
services on smart
-
phone
using


GPS/Galileo/GNSS and GIS

2


2


2

2

2

(2)
Optimization

approach in
WDM
/GMPLS

optical
networ
ks
for

design/control/management

(3) Desktop grids based on
v
olunteer

computing

(4) QoS ressource allocation in IP over WDM networks

(5) Open
-
source GIS (Geographic Information System)

Truong thi Dieu Linh, PhD

(1) Optical Network design: LTD an
d RWA problem

1

1

1

1

1

(2) Survivable optical network problem

(3) VoIP application on Mobile handset

(4) Developing application on GIS

(5) Internet over Bluetooth

(Form
-
1)

Dr. Pham Huy Hoang

(1)
Research
on

standard process of
various

project manageme
nt
(general project, software development project, offshore project for
Japanese customer, network installation project, system integration
project, etc…)
.
This theme includes

creat
ion of

a
p
ortal of
p
rojects
m
anagement

2




2



2

(2)
Research
on

the con
cept of “hardware abstraction” and “Cloud
Computing”

(
install an environment of “Cloud Computing” based on
open source products. Finally, create some add
-
on to this system
)

(3)
Research
on

Web
s
ervice mechanism
(including c
reat
ion of

information system
adapted to handheld mobile devices
)

Dr. Nguyen Linh Giang

(1) Quality of service of Multimedia Communication

1

2

2

(2) Network intrusion detection by anomaly detection

(3) Multimodal Biometric Authentication

As
soc Pr
of
. Dang Van Chuyet

(1) Data Co
mpression.

Study and Implement. evaluate


some algorithmes of Image or Audio
Compression

2

Assoc. Prof.

Nguyen Thi Hoang Lan


(together with: MSc.
Tran Quang
Duc
,
Eng. Tran Nguyen Ngoc, Eng.
Bui Trong Tung)

(1)

Developing VoIP application using Open Source H
.323
(co
-
supervisor: Tran Quang Duc)

(2)

Design & implement VoIP application (PC
-
to
-
PC model)
(co
-
supervisor: Tran Quang Duc)

(3)

Authentication based on
Biometric
Public Key Infrastructure
(
BioPKI
) (co
-
supervisor: Bui Trong Tung & Tran Nguyen Ngoc)

1


1


1

(Form
-
1)


Sup
ervisor Name

Theme Titles

Maximum number of
students

Dr. Ngo Quynh Thu

(1)
Quality of Service for Sensor Networks

Wireless Sensor Networks consists of spatially distributed sensors
to cooperatively monitor physical or environmental conditions such
as temp
erature, sound, vibration pressure…The development of
wireless sensor networks was motivated by military applications, or
environmental applications…

Quality of Service for Wireless Sensors Networks will focus on
providing different service level for packe
ts delivered within this
networks

(2)
Voice and Video over Wireless LAN IEEE 802.11

Wireless local area network have become popular in the home due
to ease of installation and the increasing popularity of laptop
computers. The transmission of different app
lications such as voice
or video over this networks will be discussed and simulated by
different simulation tools (NS2, OPNET…)

(3)
UMTS (Universal Mobile Telecommunications System)

UMTS is one the third generation mobile telecommunications
technologies wh
ich is also being developed in to a 4G technology.
During the GR time, students will focus on the main technology of
1









1





1


(Form
-
1)

this systems, and evaluate its performance by using different
network simulation tools

Dr. Le Thanh Huong


(1) Automati
cally gathering and categorize webpages

This system can be considered as the first module for the information
extracting system mentioned in (2). The purpose of this system is to
collect all webpages on a specific domain. In order to do that, we can
search

in the Internet for such webpages by using certain keywords
and then categorizing searching results into two groups: the webpages
that we are really looking for and the ones that we aren’t. Techniques
to find such keywords, to collect data and to categori
ze webpages are
the main forcus in this research.

1

(2) Information Extraction (IE)

The purpose of this research is to implement a system that
automatically extracts information (such as name, telephone, etc.) from
webpages and exports these values to a

relational database. Rule
-
based
approach and
statistical approach are considered in this research. Each
approach requires one student to be involved.





2

(3) Semantic Web Application

The purpose of this research is to implement a system that automatic

generat
ion of
a semantic web

(e.g., similar to
http://semanticweb.org/wiki/Main_Page

). T
he input of the system is a
relational database, which is extracted from the system in (2). Three
1

(Form
-
1)

m
ain tasks of this research are:

-

Convert from the above mentioned database to an appropriate
data format that can be used in a semantic web application;

-

Propose a method to organize the web page in a most
convenient way for users;

-

Propose a method to build

a semantic network in order to
control semantic relations among concepts in the implemented
system.

(4) Question Answering System

-

Research approaches in question analysis and answer generation in a
specific application domain (such as museum or tour
ism).

-

Implement a proposed approach. The input of the system is a
question posed by human. The output is an appropriate answer, which
often corresponds to an attribute’s value in a database record.

-

Question types

: questions about time (when), place
(where), author
(who), content (describe, what is)

2

(5) Anaphora resolution in dialogue (will be integrated with the
system in (4) )

In a dialogue, pronouns are often used to replace noun phrases or verb
phrases that are mentioned in previous sentences.

These noun phrases
or verb phrases sometimes are even ignored in the dialogue (ellipsis
phenomenon). In order to make a system understand correctly
1

(Form
-
1)

questions posed by human, we need a mechanism to clarify questions
by replacing pronouns or ellipsis by the

corresponding noun phrases
or verb phrases. This is the anaphora resolution problem in dialogue.

Assoc.Prof. Tran Dinh Khang

(1) Fuzzy logic and approximate reasoning

-

study some aspects of fuzzy sets, fuzzy reasoning, fuzzy systems …
to process vagu
e, incomplete and inexact information

-

apply approximate reasoning in different problems

3



(2) Some methods for solving decision making problems

-

study some aspects of decision making (DM), as multi
-
attribute,
multi
-
criteria, multi
-
objective DM

-

som
e methods for solving DM problems

-

design and implement applications

2

(3) Linguistic variables and computing with words

-

study of linguistic variables, linguistic values, hedge algebras …

-

structure of linguistic values and semantic representation

-

applications in knowledge representation and processing

2

Dr. Ta Tuan Anh

(1) Open sources for building portals with SOA

+ Study SOA, Web service, EAI, ESB,...

+ Open source portals such as Liferay, uPortal,...

+ Applications: University portal, e
-
governm
ent,...

2 students for each theme

(2) Image processing for camera surveillance

+ Motion detect and recognition of vehicles on route

(Form
-
1)

+ Number plate recognition

(3) Digital video processing for IP cameras

+ Digital video recorder

+ Video conferencing

+ Video wall for camera monitoring

(4) Embedded systems for location
-
aware services

+ GPS and GPRS devices

+ GIS applications: vehicle monitoring, taxi/bus control center

(5) Ontology
-
based content processing for internet data

+ Semantic web techno
logies

+ Information extraction

+ Question and answer systems

+ Social network

(6) Ontology
-
based knowledge systems

+ Representation of expert know
-
how under ontology

+ Case based reasoning with ontology

+ Applications: E
-
learning, E
-
health

Dr. Nguye
n Kim Anh

(1) Constructing a Natural Language Interface to Relational Databases

A natural language interface to relational databases accepts natural
language questions as inputs and generates answers under the form of
tables or short answers.

2

(2) (2)

Query relaxation in a cooperative database system

2

(Form
-
1)

Query relaxation is a query refining process to produce answers that might be a
superset of the expected answers.

(3) Data summarization in a cooperative database system

Data summarization is a learnin
g process to produce general descriptions or
overall statistics from detailed databases.

2

(4)
Providing
inten
t
ional

answers in a cooperative database system

I
nten
t
ional

answers
are created based on
generalization of

concept
s or rules that
better summari
ze the given answer set

2

(5) Providing approximate answers in a cooperative database system

Direct data retrieval may not always find enough answers for a user.
Furthermore, a user may like to know more information than the direct answers to
query for d
ecision making. Providing approximate answers is mechanism to
provide associated or neighborhood answers to a query.

2

Dr. Nguyen Huu Duc

(1)

Cryptanalysis on GPU
-
based desktop supercomputer

1

1

1


1

1

(2)

Multi sequence alignment on GPU
-
based desktop supercomp
uter

(3)

Constructing graduation thesis database for joint libraries based
on data grid technology

(4)

Detecting plagiarism in graduation thesis

(3)

Constructing grid services for solving bio
-
informatics problems

Dr. V
u

Tuy
e
t Trinh



(1) Tourism information

portal

This project can be developed in some directions:

-

design & implement the portal

3

(Form
-
1)

-

information integration (for the portal)

-

context
-
based information services (user’s profile, location, … )

(2) Scientific search engine

This project can be

developed in some directions:

-
Design & implement scientific document management system
-
information extraction

-
search techniques: keyword
-
based search (in documents and in
relational database), references search, …

2

(3) Mobile programming

This pr
oject can be developed in some directions:

-
Design & implement application for mobile devices

-
Information service adaptation for mobile (may consider services in
(1), (2) or others)

2

Dr NGUYEN Nhat Quang

(1)
Push mobile recommendation

This research p
roblem relates to the provision of mobile push
recommendations. Most existing recommender systems provide
recommendations to users in a pull
-
delivery manner, where users
explicitly request for some product or service recommendations.
However, in many pract
ical application problems, the available
products and services change quickly and often, pull
-
delivery
recommendations seem less effective in helping users keep track of
their interested products and services. To provide push
-
delivery
1

(Form
-
1)

recommendations, the

system must decide: who will receive the
recommendations, what recommendations should be (automatically)
pushed to the user, when the recommendations should be pushed to
the user, and how to visualize the push recommendations on the user’s
mobile device.
In this research project, we want to define a mobile
push recommendation method that exploits both the user preferences
model and the user mobility model to compute a (small and potential)
set of push
-
delivery recommendations.

(2)
Personalized recommen
dations for groups of mobile users

Most existing recommender systems are designed to provide relevant
recommendations for individual users. The goal of the project is to
design a recommendation methodology for providing personalized
recommendations to a gr
oup of mobile users. For example, a family
traveling to a city wants to find an appropriate accommodation, or a
group of students want to find suitable learning materials for their
course project work. Different from an individual
-
user
recommendation probl
em, in the group recommendation problem the
system must elicit, integrate and exploit the needs and preferences of
all the users of the group in the recommendation process. Moreover, in
many cases there are conflicts in the preferences of the users in the
group. In such case, the system should support the group to solve the
occurred conflicts.

1

(Form
-
1)

(3)
Repairing empty
-
result queries in mobile recommender
systems

In recommender systems that represent user preferences by logical
constraints, the execution of a

logical query may return no items. Such
a logical query is called a failing (over
-
constrained) query. Given a
failing query, the system should repair it to get a new one whose
execution returns a non
-
empty result set. The motivation of the failing
query r
epair is that users usually dislike to get back from the execution
of a query such a message: "no products found". The goal of this
research project is to develop a failing
-
query repair methodology
appropriate for mobile recommender systems. The designed
m
ethodology should exploit the knowledge contained in 1) the user's
query and 2) the past recommendation sessions. The proposed
methodology should take into account the limitations of mobile
devices and the behavior of mobile users.

1

(4)
Maintenance of t
he case base in case
-
based mobile
recommender systems

This research problem relates to the maintenance of the system’s
knowledge base in case
-
based mobile recommender systems (RSs).
Case
-
based RSs are knowledge
-
based RSs that exploit Case
-
Based
Reasoning (
CBR)


a machine learning approach


to generate
personalized recommendations. A case
-
based RS maintains a set of
1

(Form
-
1)

cases, i.e., a case base that stores the system’s knowledge of
previously solved recommendation problems and their solutions. CBR
is based on
learning from previous experiences. In existing case
-
based
RSs, when a recommendation session finishes it is retained as a case
in the system’s case base, but all the solved problems are considered
as equal in the problem
-
solving quality. Furthermore, exis
ting
case
-
based RSs treat all cases in the case base as equally important in
the case retrieval process. In this research project, we want to
associate each solved (mobile recommendation) case with a quality
measure value, and define a method for exploitin
g the case quality
measure value in the case retrieval process.

(5)
Website personalization for mobile users

Website personalization aims at adapting the behavior of a system in
order to deliver contents and services which are best suited to an
individ
ual user’s needs and preferences in a certain usage situation.
Users often want to find their interested information by means of an
easy and quick navigation, to see the information in their favorite
layout, and to get the services related to the informati
on in an efficient
manner. Personalization in mobile environment (i.e., when users use
some kind of mobile devices to access the website) aims at providing
information and services personalized for a particular mobile user.
Unfortunately, most websites tod
ay are optimized exclusively for
1

(Form
-
1)

desktop users, and poorly serve mobile ones. The goal of this research
project is to develop an appropriate methodology of (links and/or
contents) personalization for mobile users that overcomes the intrinsic
limitations of

mobile usage environments and suits the behavioral
characteristics of mobile users.

(6)
Mobile news recommendation

There exist a large number of websites that provide daily news (e.g.,
in Vietnamese or in English) in very diverse topics/categories. T
he
goal of this research project is to develop an appropriate news
recommendation methodology to provide interested news for mobile
users. The proposed methodology, based on machine learning
approaches, must exploit the system’s knowledge of user’s needs a
nd
preferences and take into account the limitations of mobile devices.

1

(7)
Email spam filtering

For many persons, everyday they receives a (large) number of
incoming emails that they have to consider; and a number of them are
spam (unsolicited) emails

for the user. Of course, everyone wants to
filter out such spam emails. However, whether an email is a spam one
depends on not only the content of the email but also the subjective
appraisal of the current user (e.g., a new camera model advertisement
emai
l can be considered as a spam one for many persons, but is not for
some ones who is interested, or is looking for, new camera models).
1

(Form
-
1)

Hence, the goal of this research project is to develop an appropriate
methodology, based on machine learning approaches,
that is capable
of accurately filtering spam emails out of a given user’s mailbox.

(8)
Loan application risk estimation

Accurate estimation of credit applications is one of the important tasks
in financial loan provision organizations (e.g., banks). Gi
ven a credit
application, the financial loan provision organization needs to estimate
the risk level of the credit application (e.g., the degree to which the
loan will be returned). The goal of this research project is to develop
an appropriate methodology
, based on machine learning approaches,
for accurately estimating credit applications.

1

Dr. Tran Duc Khanh

(1) Machine Learning Approach to Domain
-
specific Web Crawler

2

1


2

2

1

(2) Document Classification and Application in Domain
-

specific
Searh Eng
ine

(3) Model Checking and Theorem Proving for Software Verification

(4) Linguistic Logic Reasoning

(5) Access Control in Computer Security

Assoc. Prof. Nguyen Thanh Thuy

(1) Development of Hybrid intelligent systems and application

1

(2) Kno
wledge base systems

1

(3) Probabilistic Reasoning in F
-
rule based systems

1

(Form
-
1)

(4) Information retrieval

1

(5) Grid/Cloud computing infrastructure in business

1

Dr. Hoang Minh Thuc

(1) Design and implementing information systems, distributed systems

1

Assoc. Prof. Huynh Quyet Thang
,





(

together with:


Msc. Nguyen Thi Thu Trang, Eng.
Hoang Anh Viet, Eng. Nguyen Tien
Thanh, Eng. Nguyen Manh Tuan,
Msc. Le Quoc,
Dr. Tran Minh (MIC)

Dr. Nguyen Viet Hai (eDT

….)


(1) BPEL (
Web Services
Business Process
Execution Language) and
Application in
Service
-

Oriented Software Metrics and Verification

5
-
10

(2) Software Complexity Measurement and evaluation based on
design specification

(3) Design monitoring and control systems using the
O
p
en
Connectivity v
ia open standards Unified Architecture

(4) Some problem in IT Governance and Business
Intellige
n
ce

(such
as Risk Management,
Advanced

Methodology in Project
Management.

(5) Cloud computing and Map

R
educe Algorithm: Architecture,
Platforms, Experime
nting, Evaluation, Developing Application

(6) Source Code Analysis and Optimization Methods and Secure
Coding Techniques

(7)
Software Verification and Testing Methodology using Design by
Contract
a
nd Eiffel

(8) Fault
-
Tolerant Software Development
and Quality Evaluation
using Markov Model

(Form
-
1)

(9) XML Documents Matching Algorithm and Application in
Sema
n
tic Web and Object Searching Application

(10) SOA and application in developing solution for E
-
Gover
n
ment

Dr. NGUYEN Khanh Van

(1)

Designing efficien
t algorithms for wireless networks: data
-
centric
storage, routing and other related topics

1
-
3 student per each of
theme 1
-
3; only one per
theme 4 or 5; 5 in total

(2)

Network Security: web vulnerabilities, DDOS attacks; analyzing
the problems, providing sol
utions and designing support software
tools.

(3)

Designing coin
-
based e
-
payment systems for being used in
Vietnam, suitable for low
-
value purchases and shopping habits of
popular Vietnamese customers.

(4)

Peer
-
to
-
peer networking: distributed hash tables and

applications

(4)

Load
-
balancing scheduling and scalability in managing sizable
clusters of
web

servers

Dr. Vũ Th


Hương Giang

(1) Developing a cloud application using SOA (service
-
oriented
architecture) approach

1
-
2 stud
ents

p
er

th
eme

(2) Monitoring and Preventing the system security incidents using
Honeynet

(3) Developing a service composition fr
amework; experimenting on
the web site for the blind (co
-
supervisor: Nguy

n Th


Thu Trang)

(4) Applying WCAG 2.0 (web content accessibility guideline) to
ensure the accessibility of composite services (co
-
supervisor:
(Form
-
1)

Nguy

n Th


Thu Trang)

(5) De
tecting attacks by collecting and analyzing related information
from log files

Assoc.Prof. Nguyen Duc Nghia

Meta
-
heuristic
s

for som
e NP
-
hard combinatorial problems

5

Dr. NGUYEN Hong Quang

(1) Tone recognition for Vietnamese speech

1

(2) Fundamental f
requency identification of Vietnamese speech signal
verifying by EGG signal

1

(3) Constructing multi
-
syllabic language model for Vietnamese speech
recognition system

2

(4)
Develop
ment of

tri
-
gram language model for Vietnamese speech
recognition system

1

(5) Integrating tone information into Vietnamese speech recognition
system

1

(6) Verifying different lexicons for Vietnamese speech recognition

1

(7) Spoken term detection for Vietnamese speech

2

Dr. Nguyen Kim Khanh


(

together with:

MSc. Bui Quo
c
Anh,
Eng. Tran Tuan Vinh, Eng.
Pham Van Thuan,

Eng. Luong Anh
Hoang, Eng. Pham Ngoc Hung)


(1)

Smart Space (Office, Lab, House...)

(2)

Mobile device networks over bluetooth

(3)

Design w
ireless sensor networks

(co
-
supervisor: Bui Quoc Anh)

(4)

Routing protcol for wireless sensor networks (co
-
supervisor: Pham
Van Thuan)

(5)

Tracking mobile station in motion based on GSM network

(co
-
supervisor: Tran Tuan Vinh)

2

1

1

1


1


(Form
-
1)

(6)

Tracking mobile position based on GSM/GPRS/3G network
(co
-
supervisor: Tran Tuan Vinh)

(7)

Increa
se the accuracy of GPS navigation based on inertial
navigation system (co
-
supervisor: Tran Tuan Vinh)

(8)

Security for wireless sensor networks (co
-
supervisor: Luong Anh
Hoang)

1


1


1


Dr. Trinh Van Loan


(1)

Speech Recognition in Real Time using dsP
IC from Microchips


(2)

Localisation of a sound source


3

2

Confirmed by GM
-
H: Date: / /